You are on page 1of 15

SPE 128359

Statistical Diagnosis(VEMST) of Flow Regime: Alternative to Pressure


Derivative Approach in Pressure Transient Analysis—Part I

Biu Victor. T. Total E & P Nigeria, Biu Emmanuel .O. University Of Port Harcourt, Onyekonwu Mike.O,
Laser Engineering Consultant

Copyright 2009, Society of Petroleum Engineers Inc.


Introduction
This paper was prepared for presentation at the 33rd Annual SPE International In the oil and gas industry, accurate reserve
Technical Conference and Exhibition in Abuja, Nigeria, August 3-5, 2009.
estimate, well capacity and efficient reservoir
This paper was selected for presentation by an SPE Program Committee following
review of information contained in an abstract submitted by the author(s). Contents of management in field development depend on
the paper, as presented, have not been reviewed by the Society of Petroleum
Engineers and are subject to correction by the author(s). The material, as presented,
precise analysis and result of reservoir monitoring
does not necessarily reflect any position of the Society of Petroleum Engineers, its techniques such as pressure transient analysis. This
officers, or members. Papers presented at SPE meetings are subject to publication
review by Editorial Committees of the Society of Petroleum Engineers. Electronic prompted the introduction of the fluid flow diffusivity
reproduction, distribution, or storage of any part of this paper for commercial purposes
without the written consent of the Society of Petroleum Engineers is prohibited.
model using only one fraction of data by Muskat
Permission to reproduce in print is restricted to an abstract of not more than 300 words; (1934) and Theis (1935). However in 1949, Van
illustrations may not be copied. The abstract must contain conspicuous
acknowledgement of where and by whom the paper was presented. Write Librarian, Everdingen and Hurst simplified the complex model
SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435. using Laplace transformation in flow problem and
published solutions of the diffusivity equation which
Abstract is the basis of pressure transient analysis.
Before the early eighties, identification of Since then, more work has been done by
flow regime has been a difficult task for reservoir renowned researchers such as MDH et al (1950)
engineer and welltest analyst until the emergence of introducing the pressure-time curve for welltest
the derivative approach. This approach has helped analysis. Horner (1951) modified the method using the
to reduce the uncertainties of the interpretation of cumulative production time concept for wells with
welltest result because key regions of radial flow and production history before shut-in for pressure build-up.
boundary features required for reservoir In 1954, MBH et al also worked on average pressure
characterization, description and evaluation have determination using the pressure-time curves.
been adequately diagnose. However the approach is However for quick well and reservoir model
complex for non-mathematician and an alternative description, an easier and faster identification of flow
method for improving its interpretation and reducing regimes is required. This prompted Agarwal et al
the difficulty of its practical application haven’t been (1970) to introduced the dimensionless type curve
discovered most especially where there is using pressure function which helps to reduce the
inconsistency in data sampling. degree of uncertainties in the estimation and
The statistical approach(VEMST) utilized interpretation of the pressure-time result. In 1979,
simple statistical tools such as StatDiff, StatDev and Gringarten et al and Ramey et al (1982) proposed and
StatExp derived from time series analysis to identify generated various dimensionless type curve for
possible unseen features, diagnose key flow regime different well and reservoir conditions.
for reservoir description and act as However the greatest breakthrough in
checkmate/alternative to the derivative approach to welltest analysis was the introduction of the
interpret complex features. derivative approach by French Mathematician
Result from channel sand, low and high Dominique Bourdet in 1983 which has been used to
permeability reservoir, infinite conductivity fractures, identified flow regimes, boundary response and
dual porosity and/or permeability reservoir show diagnose complex features.
same trend as the derivative approach and also
diagnose unseen features such as dual permeability,
geologically interpreted as layered reservoir system
and also double porosity model with high degree of
accuracy.
2 T. Biu, E.O. Biu and M.O. Onyekonwu SPE 128359

Radial Flow Spherical Flow

Linear Flow Bilinear Flow

Fig. 1 - Distinct Flow Patterns In Transient Tests (Ehlig-Economides


et al., 1994

The Derivatives Approach


Fig. 2 - Type Curves For A Well With Wellbore Storage And
The diagnosis of flow regimes in a pressure Skin Effects (Bourdet et al., 1983)
transient analysis, which appear as distinctive
patterns in the pressure-derivative curve, is a vital It is observed that at early times all the curves merge
point in welltest analysis since each flow regime into a unit-slope straight line corresponding to pure
reflects the geometry of the flow streamlines in the wellbore storage; however the derivative curves tend
tested formation. Hence for each flow regime to a single horizontal line, representing pure radial
identified, a set of well and/or reservoir parameters flow at late times. Several sets of type curves have
can be estimated using the region of the transient been developed to account for different
data that exhibits the characteristic pattern combinations of wellbore, reservoir characteristics,
behaviour. As shown in Figure 1, there are a number boundary effects associated flow regimes, and also
of flow regimes that exhibit characteristic flow computation of well and reservoir parameters. This
patterns, such as radial, spherical, linear, bilinear, demonstrates that the log-log plot of the pressure-
etc., which appear as distinct fingerprints in the derivative is a powerful tool for model identification
pressure-derivative curve, then such region of in pressure transient analysis.
pressure response will be analyzed in the
specialized plot. The algorithm for the pressure derivative model is
show below;
In addition to single diagnostic tools and plots, type
curves were also developed in order to help identify
⎛ ∂p ⎞ ⎛ ∂p ⎞
the model selection and the correct curves to be ⎜ ⎟ = t ⎜⎜ ⎟⎟
used in the specialized analyses. Figure 2 shows ⎝ ∂ ln t ⎠ ⎝ ∂p ⎠ i
series of type curves for different values of skin ⎛t t ⎞ ⎛t ⎞
ln⎛⎜ i ⎞⎟Δpi+j ln⎜⎜ i+j i−k 2 ⎟⎟Δpi
t ln⎜ i+j ⎟Δpi+k
factor and wellbore storage constant, considering ti ⎠
i ⎠
⎝ i−k ⎠ + ⎝ − ⎝
t t
infinite acting reservoir behaviour. =
⎛ ⎞ ⎛ ⎞ ⎛ ⎞ t ⎛t ⎞
ln⎜ i+j ⎟ln⎜ i+j ⎟ ln⎜ i+j ⎟ln⎛⎜ i ⎞⎟ ln⎛⎜ i ⎞⎟ln⎜ i+j ⎟
t t t t
⎝ ti ⎠ ⎝ ti−k ⎠ ⎝ ti ⎠ ⎝ ti−k ⎠ ⎝ ti−k ⎠ ⎝ ti−k ⎠
………………………………………………………….1
Horne, R.N.(1995).

However the algorithm is difficult for practical


application by non-mathematician with strong
welltest analytical and interpretation skills. Hence
there is need to introduced a simplistic approach to
reduced the difficulty in analyzing pressure transient
test and improve the interpretation of complex and
unseen features where there is inconsistency
sampling data.
3 Statistical Diagnosis (VEMST) of Flow Regime SPE 128359

The Statistical Approach


The statistical models are derived from differencing
method in time series analysis which is common in 1st Pressure Diff
Fixed Initial Value
advanced statistical forecasting where it is often
used to transform a nonstationary time series into a
stationary time series. Intuitively, a time series is
stationary if the statistical properties such as mean Difference Divided Difference Divided By SSE
and variance of the time series are essentially (StatDiv)Fixed Initial Value (StatSSE)Fixed Initial Value
constant through time.
If we have observed n values y1,y2,………….yn of a
time series, we can use a plot of these values
2nd Pressure Diff Time Diff Divided
(against time) to help us determine whether the time
series is stationary. If the n values seem to fluctuate
with constant variation around a constant mean,
Standdev SQRT (STDEV, SSE)
then is reasonable to believe that the time series is
stationary. However if the n values do not fluctuate
around a constant mean or do not fluctuate with
constant variation, then it is reasonable to believe Product--StatDev Exponential--StatExp
(Statdiv, SQRT, 2nd Diff Pressure) (Statdiv, SQRT, 2nd Diff Pressure)
that the time series is non-stationary.
The first difference of the time series values
y1,y2,…….yn are; Fig3 Workflow For Statistical Models Formulation
zt = yt − yt −1
………………………………………….2
Where t=2...n
To illustrate, we list the original values and the first
difference values of a time series Statdiv
Statdiv
Original values First difference Statdiff
y1 (Semilog)
Statsse
y2 z2 = y2 − y1
y3 z3 = y3 − y2
yn-1
yn zn = yn − yn −1
Although in statistical data analysis for time series
and forecasting, first difference may not accurately Statdev Statexp
transform the nonstationary time series into (Log-log) (Log-log)
stationary time series, hence the second difference
approach serve as an alternative to such condition

The statistical approach (VEMST) utilized simple Fig4 Statistical Models For Flow Regime Identification
statistical tools such as the product and exponential
of 1st and 2nd difference of well pressure tied to the
standard deviation and sum of square difference of
this data to generate the statistical models such as The VEMST algorithm for StatDiff is shown below;
StatDiff, StatDev and StatExp which identify possible
unseen features, key flow regimes for reservoir Z pi = P0 − Pi ...................................................3
description and improve interpretation of complex
features. Figure 3 shows the statistical workflow for Z t i = t0 − ti .......... .......... .......... .......... .......... ...4
the model formulation while Figure 4 displays the Z pi
models generated using the approach. The idea is to Statdiv = ................................................5
keep it simple for non-mathematician with strong P1
reservoir and welltest analytical skills. SSE
⎡Zp ⎤
StatSSE = ⎢ i ⎥ .......................................6
⎣ P1 ⎦
4 T. Biu, E.O. Biu and M.O. Onyekonwu SPE 128359

Where the SSE= Sum of square error for the Statdiv same late time response as StatExp. To improve on
and Equation3. The StatDiff comprises of StatDiv the interpretation of this unseen feature in the
and StatSSE. derivative, the StatDiff was plotted as seen in Figure
For the VEMST diagnostic models such as StatDev 6b which shows three flow regimes (radial—
and StatExp, the algorithm is shown below; pseudosteady--radial) confirming the features
diagnose by the StatExp and StatDev. Integration of
⎛ ⎞ ⎛ ⎞ ⎛ ⎞ petrophysics and geology will aid a better
Statdev= Statdiv*⎜⎜ Z p − Z p ⎟⎟ * stdv⎜⎜ Z p − Z p ⎟⎟ *⎜⎜ Zt − Zt ⎟⎟ * SSE interpretation of this unseen feature by the
⎝ i i −1⎠ ⎝ i i −1⎠ ⎝ i i −1⎠
.......................................................................................7 derivative. This is expected in part II of this paper.

StatExp = Statdiv * Exp(− Z pi − Z pi −1 ( ) − sdev


) Channel Sand Reservoir System
.....................................................................8a A build-up test after a period of 720 hours flow with
Where permeability about 340md and reservoir pressure of
( )(
Sdev= stdvZpi − Zpi −1 * Zti − Zti −1 * SSE ) 2900 psi.

.......................................................................................8b Result

The generated statistical models in Equation 5, 6, 7 In Figure 7a, the StatDev late time response shows
and 8a are plotted against time in a log-log and a long period of linear flow, which is characteristic of
semi-log graph as indicated in Figure 4. However flow in channels. This channel feature is also
the Statdev and StatExp are diagnostic in nature depicted in the derivative as indicated in Figure 7b at
exhibiting same trend as the pressure derivative plot late time response. In addition, Figure 8 shows the
while the Statdiff is more of the pressure-time curve radial-linear flow response sharply depicted by
but differ in that flow regime can also be easily StatSSE which is in-line with the reservoir and
identified from the plot. Nevertheless several case boundary response diagnose by the StatDev.
studies on different reservoir system are
Infinite Conductivity Fracture Reservoir System;
comprehensively discussed to show the efficacy of
the approach. For an infinite conductivity fracture reservoir as show
in Figure 9a, the StatDev late time response shows
a long period of linear flow, characteristic of flow in
Discussion of Case Studies infinite conductivity fractures. The derivative in
Figure 9b exhibits the same pressure trend.
All pressure data used for the case studies are Nevertheless, in Figure 10, the StatDiff (most
design data (test and examples) obtained from especially StatSSE) shows a clear feature of infinite
Automate Windowstm application welltest demo conductivity fracture with a long linear flow regime
software (1995) and Onyekonwu (1997) example. validating the reservoir response identified by the
However this paper has three parts in which part I StatDev.
focuses only on the diagnostic nature of the
statistical tool as depicted in the case studies. Dual Porosity Reservoir System

High K Reservoir In Niger Delta Nigeria In Figure 11a, the StatDev identified key features for
the dual porosity model including, radial flow within
The reservoir permeability ranges between 800md the primary and secondary porosity. The dip
and above, with light oil PVT properties. The well between the radial flows along the porosity model
capacity is above 1000 STB/Days with reservoir confirmed the dual porosity model as indicated in the
pressure close to 5000 psi. StatDev plot. The derivative in Figures 11b exhibits
same pressure trend. In addition, the StatDiff depicts
Result double radial flow (primary and secondary porosity
response) as seen in the semilog- pressure-curve
The StatExp diagnostic plot as seen in Figure 5a
response of the dual porosity model in Figure 12.
depicts the radial flow regime at transient condition
and pseudo-steady flow at the late time. However Low Permeability Reservoir System;
the derivative plot in Figure 5b exhibit the same
pressure trend but differ after the late time response Infinite Acting Reservoir
in which StatExp indicates the present of another
radial flow geologically interpreted as a double This is a drawdown test in a well with wellbore
permeability feature, possibly a layer reservoir storage and skin. In Figure 13a, the StatDev
system. In Figure 6a, the StatDev also exhibits the indicates an infinite acting reservoir response. The
5 Statistical Diagnosis (VEMST) of Flow Regime SPE 128359

derivative shows same feature as seen in Figure Conclusion


13b. However, in Figure 14 the StatDiff indicates the
radial flow infinite acting in nature as depicted by The following inference was drawn after analysis
StatDev. and interpretation of the result of case studies
utilizing the StatDev and StatDiff approach:
Close Boundary Reservoir
Both approach exhibits same well, reservoir and
This is a drawdown tests showing the effect of a boundary response as seen in all cases considered.
closed boundary response in the reservoir. In Figure Also possible flow regimes are easily identified.
15a, the StatDev depicts a radial flow regime with a
unit slope straight line at late time indicating a close StatDev and StatExp identified unseen features
boundary response. The derivative also exhibits such as dual permeability response, geologically
same radial and boundary response in Figure15 b. interpreted as layered reservoir system in the 1st
The Statdiff plot as seen in Figure 16 shows two flow case considered in the derivative at late time
regimes (radial-pseudosteady) confirming the response and its interpretation confirmed by StatDiff.
reservoir and boundary response diagnose by the The radial-pseudosteady-radial response of StatDev
StatDev. Nevertheless noisy data was noticed within and StatExp make the VEMST approach a detail
the radial region of both StatDev and the derivative. diagnostic tool.
Acidized Reservoir System; Late time boundary response for cases with
limited data can be diagnose as indicated in Low
This is a build-up test showing the reservoir and High permeability reservoir system where there
response in a stimulated well. In Figure 17a, the is sharp change in pressure trend toward the end of
StatDev shows a flat radial flow regime indicating a the data.
fracture well model /or acidized reservoir response.
The derivative exhibits same pressure response as The Statistical approach (StatDev, StatExp and
seen in Figure 17b. Nevertheless, in Figure 18, the StatDiff) are precise in identification of flow regimes
StatDiff (especially the StatSSE) depicts an infinite and reservoir features as exhibited by the derivative
radial flow regime which is a clear feature of a approach.
stimulated/or fracture model response validating the
reservoir response diagnose by the StatDev However, the approach is simple, easily
applicable by non mathematicians as indicated in
Superposition Time Effect; the workflow.
This is a build-up test to show the superposition time Good statistical reservoir diagnostic tool with
effect on reservoir response. In this case, the well high degree of accuracy, as it’s identify same
producing time of 21.6 hrs is short relative to the reservoir features with the derivative approach.
shut-in time, so the effect of time superposition is
seen on the StatDev and derivative plot. In Figure Future Work
19a and 19b, the derivative and the StatDev exhibit
similar pressure trend depicting the superposition However to advanced its application in the industry,
time effect on reservoir response. However the more work including the following is expected
StatDev differs slightly due to sharp dropped in
pressure in the plot. This feature can be easily Integrate geological and petrophysical studies to
diagnose and interpreted using other reservoir new and/or complex features diagnose for better
analysis tools. interpretation

High Permeability Reservoir System, Horizontal Well Testing the approach in more complex well and
reservoir model, and gas fields data.
This is a drawdown test to show the reservoir and
boundary response for horizontal well in a high K Introduction of slope/or statistical parameter to
reservoir system. In Figure 21 and 22, the StatDev identify well and reservoir model
and derivative exhibit same pressure trend,
indicating an early time and late time radial flow Reduction of noise in some data using statistical
behaviour interpreted as a high permeability smoothing factor, however the derivative also exhibit
reservoir with horizontal well. There is some such noise for all cases considered.
indication of hemiradial flow in this response in the
derivative. However the StatDev exhibits similar Estimation of permeability from the approach
trend and reservoir features.
6 T. Biu, E.O. Biu and M.O. Onyekonwu SPE 128359

Reference: Horne, R.N. (1995). Modern welltest analysis.


Petroway Inc, Palo Alto, California.
Anderson T.W (1971),”The statistics analysis of time
series”, New York wiley Horner. D. R., (1951). Pressure build-ups in wells.
PROC. Third world Pet-congs E.L. Brill. Leiden
Agarwal, R.G., AL-Hussainy, R., and Ramey, H.J.JR II.503-521. Also reprint series, No 9 Pressure
(1970). An investigation of wellbore and skin effect analysis methods, SPE of AIME. Dallas (1967) 25-
in unsteady liquid flow I. Analytical; treatment. SPE 43.
J., Sept 279-290.
Bourdet, D. et al. (1983) A new set of type curves Matthews, C.S, Brons, F. and Hazebroek, P.,(1954)
simplifies well test analysis. World Oil, 196(6), 95- A method for determination of average pressure in a
106. bounded reservoir. TRANS AIME, 201, 182-1914.
o
Also reprint series N P- Pressure analysis method,
Box.G.E.P and Jenkins G.M (1976)” Time series SPE, 51-60.
analysis, forecasting and control, 2d Ed
Sanfrancisco holden-day Miller, C.C, Dyes, A.B., and Hutchinson, C.A.,(1950).
Estimation of k and reservoir pressure from
Ehlig-Economides, C.A.; Hegeman, P.; Vik, S. bottomhole pressure buildup characteristics. TRANS
(1994) Guidelines simplify well test interpretation. Oil AIME, 189, 91-104
and Gas Journal, Jul. 1994.
Muskat, M., (1934). The flow of compressible fluid
through porous media and some problems in heat
http://www.petroway.com/amw.htm; Automate
conduction physics, march
Windowstm Application Welltest Software (1995) Test
and Examples of welltest, Demo copy” Petroway Inc, Onyekonwu, M.O (1997). General Principle of
Palo Alto, California. bottomhole pressure tests. Laser Engineering
Consultant, Port Harcourt, pp56-58
Gringarten A.C, Bourdet D.P, Landel, P.A, and
Kniazeff V.J,(1979) A comparison between different Ramey.H.J. Jr., (1982) Pressure Transient testing.
skin and wellbore storage typecurve for early time J.Pet tech, July, 1407-1413
transient analysis, SPE paper 8205, 54th Annual fall
meeting Las Vegas Texas. Theis, C.V., (1935). The relation between the
lowering of
the piezonmetric surface and the rate and duration Zheng, S.-Y. (2006) Fighting against non-unique
of discharges of well using ground-water storage solution problems in heterogeneous reservoirs
TRANS. AGU 519-524. through numerical well testing. In: SPE Asia Pacific
Oil and Gas Conference and Exhibition, Adelaide,
Van Everdigen, A.F, and Hurst, W.(1949). The Sep. 2006, SPE 100951.
application of Laplace transformation for flow
problem in reservoirs. TRANS AIME.186, 305-324. Nomenclature And Symbols

Zheng, S.-Y. (2008a) Radial flow and well testing VEMST; Victor Emmanuel Mike Statistical
basics, Well Testing lecture presentation, Institute of Techniques
Petroleum Engineering, Heriot-Watt University,
Edinburgh, Jan. 2008. Statdiff; Statistical difference
Statdev; Statistical deviation
Zheng, S.-Y. (2008b) Wellbore storage, radial
Statdev; Statistical exponential
derivative and type curve analysis, Well Testing
lecture presentation, Institute of Petroleum StatSSE; Statistical sum of square
Engineering, Heriot-Watt University, Edinburgh, Jan. Statdiv; Statistical divisible
2008. Wellbore Storage Region
Radial Flow Regime
Zheng, S.-Y. (2008c) Well testing semi-infinite acting Pseudosteady flow
reservoir systems, Well Testing lecture presentation,
Institute of Petroleum Engineering, Heriot-Watt Linear Flow
University, Edinburgh, Jan. 2008.
Appreciation; To Prof Mike Onyekonwu for his in-
depth technical support to this paper.
7 Statistical Diagnosis (VEMST) of Flow Regime SPE 128359

High K Reservoir in Niger


Delta Nigeria

100000
100
StatExp
Derivative
1000
StatExp Model For Flow Regime Diagnosis
10

10

1
0.1
Derivatives Diagnostics Of Flow Regime

0.001 0.1
0.00 0.01 0.10 1.00 10.00 100.00 0.001 0.010 0.100 1.000 10.000 100.000
Time(hrs)
Time(hrs) derivative

Fig 5a and b Comparison of StatExp and Derivative Diagnostics Approach

1 72 25

StatSSE
20
StatDev 54

15
StatDiv
0.1 36
10

18
5
StatDev Diagnostics Of Flow Regime Pdd ^sse P diff div

0.01 0 0
0.001 0.010 0.100 1.000 10.000 100.000 0.001 0.010 0.100 1.000 10.000 100.000
Time(hrs) Time(Hrs)

Fig 6 a and b StatDev And StatDiff Diagnostic Plot For High Perm System
8 T. Biu, E.O. Biu and M.O. Onyekonwu SPE 128359

Channel Sand Reservoir

10
1000
pddexp-Test2
StatDev
Derivative
1

100
0.1

Derivative Approach
0.01
10
0.001 0.010 0.100 1.000 10.000
0.001 0.010 0.100 1.000 10.000 100.000
Time hrs pddexp Time Hrs derivative

Fig7 a and b Comparison of StatDev and Derivative Diagnostics Approach

10 Test2
25
StatSSE
8 20

6 15
StatDiv
4 10

2 5

0 0
0.001 0.010 0.100 1.000 10.000
Time Hrs P diff div Pdd ^sse

Fig 8 Comparison of StatSSE & StatDiv Diagnostic Plot for Channel Sand System
9 Statistical Diagnosis (VEMST) of Flow Regime SPE 128359

Infinite Conductivity Fracture


Reservoir

1000
1000

StatDev Derivative
10 100

0.1
10
Derivative Approach
Statistical Approach--pddexp

0.001
0.01 0.1 1 10 100 1
Time Hrs pddexp 0.01 0.1 1 10 100
derivative
Time Hrs
Fig 9 a and b Comparison of StatDev and Derivative Diagnostics Approach

400 35

28
300
StatDiv
21
200
StatSSE
14

100
7
Statistical Approach---Example5
0 0
0.01 0.1 1 10 100 1000
Time Hrs
Pdd ^sse P diff div

Fig 10 Comparison of StatSSE & StatDiv Diagnostic Plot for Infinite Conductivity System
10 T. Biu, E.O. Biu and M.O. Onyekonwu SPE 128359

Dual Porosity Reservoir


System

100
1

Derivative
StatDev
0.1
10
Dip
Dip
0.01

1
0.001
Derivatives Diagnostics Of Flow Regime--Test24
Statistical Diagnostics Of Flow Regime--Test
0.0001 0.1
0.00 0.01 0.10 1.00 10.00 100.00 1000.00 0.010 0.100 1.000 10.000 100.000 1000.000
Time(hrs) Time(hrs) derivative

Fig 11 a and b Comparison of StatDev and Derivative Diagnostics Approach

12
StatDiv
6

8 StatSSE
4

4
2

0 0
0.001 0.010 0.100 1.000 10.000 100.000 1000.000
Time(Hrs)
Pdd ^sse P diff div

Fig 12 Comparison of StatSSE & StatDiv Diagnostic Plot for Low K System
11 Statistical Diagnosis (VEMST) of Flow Regime SPE 128359

Low K Reservoir,
Infinite Acting System

100
10000
StatDev
Derivative
10
1000

1 100

Statistical Diagnostics Of Flow Regime--Example1.0 Derivatives Diagnostics Of Flow Regime

0.1 10
0.01 0.10 1.00 10.00 100.00 0.001 0.010 0.100 1.000 10.000
Time(hrs) Time(hrs) derivative

Fig 13 a and b Comparison of StatDev and Derivative Diagnostics Approach

100 High-perm
25
StatDiv

20
75
StatSSE
15
50
10

25
5

0 0
0.01 0.10 1.00 10.00 100.00
Time(Hrs) Pdd ^sse P diff div

Fig 14 Comparison of StatSSE & StatDiv Diagnostic Plot for Low K System
12 T. Biu, E.O. Biu and M.O. Onyekonwu SPE 128359

Low K Reservoir,
Close Boundary System

1000 10000

StatDev
Derivative
100 1000

10 100

StatDev Diagnostics Of Flow Regime-Example1A


Derivatives Diagnostics Of Flow Regime-Example 1A
1 10
0.01 0.10 1.00 10.00 100.00 0.01 0.10 1.00 10.00 100.00
Time(hrs) Time(hrs) derivative

Fig 15 a and b Comparison of StatDev and Derivative Diagnostics Approach

160 30
StatSSE

120
StatDiv 20

80

10
40

0 0
0.01 0.10 1.00 10.00 100.00
Time(Hrs) Pdd ^sse P diff div

Fig 16 Comparison of StatSSE & StatDiv Diagnostic Plot for Low K System
13 Statistical Diagnosis (VEMST) of Flow Regime SPE 128359

Low K Reservoir,
Acidized Well

10000 1000

1000 100

100
10
StatDev
10 Derivative
1
1
0.1
0.1
Derivative Approach
StatDev Approach--Test1
0.01
0.01
0.01 0.1 1 10 100 1000
0.01 0.1 1 Time Hrs10 100 1000 Time Hrs derivative
pddexp

Fig 17 a and b Comparison of StatDev and Derivative Diagnostics Approach

150000 15000

12000
100000
StatDiv 9000
StatSSE
6000
50000

Statdiv & StatSSE Method 3000

0 0
0.001 0.01 0.1 1 10 100 1000
Time Hrs P diff div Pdd ^sse

Fig 18 Comparison of StatSSE & StatDiv Diagnostic Plot for Low K System
14 T. Biu, E.O. Biu and M.O. Onyekonwu SPE 128359

Low K Reservoir,
Superposition Time Effect

1000 10000

100
StatDev 1000 Derivative
10
100
1

0.1 10
Derivative Approach
Statistical Approach--Example2
0.01
1
0.01 0.1 1 10 100
0.01 0.1 1 10 100
Time Hrs pddexp
Time Hrs derivative

Fig19 a and b Comparison of StatDev and Derivative Diagnostics Approach

40
StatSSE

30

20

StatDiv

10

Example 2

0
0.01 0.1 1 10 100
Time Hrs Pdd ^sse P diff div

Fig21 a and b Comparison of StatSSE & StatDiv Diagnostic Plot for Low K System
15 Statistical Diagnosis (VEMST) of Flow Regime SPE 128359

High K Reservoir,
Horizontal Well

100

Derivative
10

Derivative Approach
0.1
0.0001 0.001 0.01 0.1 1 10 100
Time Hrs derivative

Fig22 Derivative Diagnostics Response For Horizontal Well With High K Reservoir

100

StatDev
10

Statistical Approach--Test6
0.1
0.0001 0.001 0.01 0.1 1 10 100
Time Hrs pddexp

Fig23 Statdev Diagnostics Response For Horizontal Well With High K Reservoir

You might also like